🗓️ Tuesday 20 May 2025  🕒 15:00-16:00

• 🟢Theme: Critically exploring data-driven technologies and practices in digital education

Exploring complex entanglements of artificial intelligence technologies and educational practice through a diffractive approach

Michael Wolfindale

Recording

Abstract

Artificial intelligence (AI) educational programmes often cover machine learning techniques shaped by specific conceptualisations of learning, a focus for my dissertation research. If one acknowledges the complexity of interrogating such entanglements of educational practice and AI technologies, and the human, nonhuman, material and abstract (more-than-human) entities involved, a methodological approach questioning traditional subject/object and human/nonhuman binaries is required. Reflexive approaches rooted in representationalism, informed by the metaphor of reflection, tend to reinforce subject/object boundaries. Haraway posits diffraction—revealing patterns produced through differences—as an alternative metaphor. A diffractive methodology, as taken forward by Barad, shifts the focus to ‘relations of difference’ within/beyond such boundaries.

‘Diffraction’ refers to the patterns produced when light, sound and other waves experience obstructions or openings. As light diffracted through an opening produces new patterns, in a diffractive analysis data is read through different theories—a transdisciplinary process enabling engagement with multiple ideas, generating tensions but new insights and questions. Such approaches provided inspiration for my dissertation research employing diffractive analysis to explore machine learning students’ conceptualisations of learning. By diffractively reading multimodal participatory conversation data through different (seemingly incompatible) theories—multimodal social-semiotics and anti-anthropocentric new materialisms—a new artefact was assembled. This artefact conveys relations between more-than-human components, through which the students’ conceptualisations of learning emerge. While behaviourist ideas—at times appearing to emphasise notions of control working against student autonomy—are made visible, a complex multi-faceted picture is presented.

During this session, I outline my diffractive methodology and data analysis process, and discuss the multimodal insights generated. Through this, I demonstrate how a diffractive analysis can critically explore entanglements of AI technologies and educational practice—from a detailed reading of data through multiple theories, new ideas and provocations emerge. Finally, I invite discussion about how such approaches might inform transdisciplinary AI education and research.